Raymond C Rosen, Jennifer L Barsky, David M Ferguson
Overview of clinical trials in female sexual dysfunction
While the conduct of clinical trials in male sexual dysfunctions (e.g., erectile dysfunction, premature ejaculation) is accepted and well established, the design and implementation of clinical trials in female sexual dysfunction is a more recent phenomenon and is not adequately defined or systematized. In contrast to the large literature on randomized, placebo- controlled trials of erectile dysfunction, which proliferated in the late 1990s and early 2000s, equivalent clinical trials in sexual arousal disorder in women have been few and far between, and have yielded much less consistent or clear-cut findings. This may be due to several factors, including differences in the relative stages of development of male and female sexual dysfunction research, as well as the complex nature of female sexual arousal. Substantial challenges also face investigators in the design of adequate clinical trials for sexual dysfunction in women. For example, large placebo effects have been noted in many studies, and discrepancies among different measures of outcome are apparent both across and within studies.
This chapter will consider the design of clinical trials for female sexual dysfunction, selection of the optimal patient population, assessment of relevant outcome measures in a standardized and clinically relevant way, and definition of treatment response criteria that are best suited to the domain of interest. The presence of multiple methodological and design issues underscores a range of problems in the conceptualization of sexual function in women, the current paradigm of clinical trials for studying female sexual dysfunction, and the current approach to diagnosis that guides much of the clinical research in female sexual dysfunction today.
The complexity of etiologic factors underlying sexual dysfunctions in women and the lack of adequate understanding of the pathophysiology or development of these disorders create a major challenge in defining a study population well suited for pharmacotherapeutic (or other) intervention. The task is complicated by culturally embedded notions of healthy female sexuality and sexual dysfunction in women,1,2 making stable conceptualizations of “sexual dysfunction” difficult to achieve. Furthermore, despite increasing evidence of the overlap in symptoms within and across sexual dysfunctions in women,3 most clinical trials of female sexual dysfunction tend to focus on a specific dysfunction, such as hypoactive desire disorder or female sexual arousal disorder. This implies that it is possible to identify a specific study population by validated diagnostic criteria. However, the artificial classification of these disorders in trials of female sexual dysfunction may pose a problem for generalizability of the study results, and may contribute to a lack of clear-cut treatment differences.
A key aspect of clinical trial design is the ability to detect a significant improvement in performance as a result of treatment. Currently, it is difficult for investigators in clinical trials of female sexual dysfunction to assess treatment outcomes and define treatment responders in a standardized way. Lack of agreement on the selection of suitable outcome measures in clinical trials of female sexual dysfunction is a major obstacle at present.4 The role of physiologic measures, such as vaginal photoplethysmography (see Chapter 10.1 of this book) or functional magnetic resonance imaging of the brain (see Chapter 10.2), is controversial and also not well established. Whereas improved physiologic function may be a more obvious goal of treatment for men (e.g., increased frequency of erections), it is unclear what role comparable physiologic measures (e.g., increased lubrication) should play as biologic markers or endpoints in a clinical trial of female sexual dysfunction.
A US Food and Drug Administration guidance document5 recommended the use of daily diary or event log frequency measures as preferable to self-administered questionnaires for assessing female sexual dysfunction outcomes in clinical trials. However, the clinical significance of behavioral frequency outcomes has been challenged.6 Recent considerations of subjective outcomes in female sexual dysfunction, including distress,7 indicate a need for increased focus on subjective issues in assessing improved sexual function in women. While there is little disagreement regarding the convergence of physiologic and subjective measures of outcome in men, available evidence suggests that physiologic and subjective measures of sexual response in women are not as strongly associated.8
Cultural and psychosocial determinations of women’s sexual function further complicate the process of defining a treatment responder in clinical trials of female sexual dysfunction. If a study includes members of a cultural or other subgroup of women who have low expectations of sexual pleasure or satisfaction, it is unclear whether or how the definition of a treatment response or successful trial outcome should be adjusted. More culturally sensitive endpoints or assessment procedures may better incorporate subjects’ expectations into the assessment of treatment outcomes, but this issue has not been addressed in recent trials of female sexual dysfunction.
Finally, the role of qualitative research in defining issues and endpoints to be addressed in clinical trials of female sexual dysfunction has not been adequately considered. In the past, qualitative and quantitative methods have been viewed as competing or rival approaches, and it was not uncommon for quantitative researchers to disregard evidence obtained from qualitative observations or interviews. More recently, however, “the classic qualitative-quantitative debate has been largely resolved with recognition that a variety of methodological approaches are needed and credible, that mixed methods can be especially valuable, and that the challenge is to appropriately match methods to questions rather than adhering to some narrow methodological orthodoxy.”9 Qualitative methods can be used specifically to assess the clinical relevance of specific measures and endpoints, and to determine whether current definitions of distress and sexual function are applicable in the clinical population.
Many of the challenges associated with conducting clinical trials for female sexual dysfunction may be attributed to the lack of a clearly defined, empirically supported model of female sexual response and satisfaction, despite recent advances in the classification of women’s sexual dysfunctions.10 The current model for clinical trials in male sexual dysfunction may not be sensitive to the complex and ill-defined scope of sexual response in women. In the post-Masters and Johnson era, it is unclear whether any one model should be used as the standard guide for diagnostic decision making. Psychosocial and cultural factors, furthermore, are not adequately measured or controlled for in these trials. Despite these limitations, the randomized, controlled trial remains the reference standard for clinical trial investigation and must necessarily be considered in detail for use in clinical trials of female sexual dysfunction.
Regulatory and medical perspectives
A variety of drugs, hormonal agents, and topical and transdermal formulations for female sexual dysfunction are currently in development, based on a broad range of hormonal and pharmacologic mechanisms (see Chapter 14.1). Along with the introduction of novel methods for the treatment of female sexual dysfunction, a wide range of new measurement approaches and assessment tools have been developed (see Chapter 11.2). These new tools have permitted more precise and reliable measurement of various aspects of sexual response in women. The use of valid and reliable outcome measures in the context of a well-controlled clinical trial is the sine qua non for medical acceptance and regulatory approval of any new drug or treatment.11
A major goal of this chapter is to review current standards of practice in the design and conduct of clinical trials in female sexual dysfunction. Many of these standards represent general concepts or principles in the design of clinical trials, although specific issues in the evaluation of new treatments for female sexual dysfunction are also addressed. These include the use of objective (physiologic) versus self-report (questionnaire) measures of sexual function, assessment of other domains of function, global patient and investigator assessments, and the use of disease-specific quality-of-life measures. Although much of the focus in recent years has been on the development of new pharmacologic or hormonal agents, similar standards of research should be applied in the evaluation of other methods of treatment, such as devices or nonpharmacologic therapies generally.
Design of clinical trials in female sexual dysfunction
The clinical trial process in most therapeutic areas, including female sexual dysfunction, is usually described as occurring in four phases (phases I, II, III, and IV), each of which is discussed below. These stages were described in detail for erectile dysfunc- tion11 and, more recently, in the Second Consultation on Sexual Medicine for female sexual dysfunction.8 Beyond the immediate goal of obtaining regulatory approval, clinical trials are intended to predict the likely risk/benefit outcomes when the drug or device enters widespread clinical use.11
These studies represent the first exposure in humans of the novel agent or device and are typically conducted in healthy individuals. The full range of phase I studies potentially relevant to female sexual dysfunction is shown in Table 16.1.1.
Initial studies in healthy volunteers typically involve administration of a single dose of the new agent. The first dose is selected on the basis of acute and chronic animal studies and, if available, data from human liver metabolism in vitro. The initial dose selected may be 30-100 times lower than the predicted threshold efficacy dose in adults. Should this be well tolerated, additional doses will usually be evaluated at regular dose increments. Escalating dose studies may employ successive naive cohorts, but frequently use the same cohort with successive doses administered between estimated washout periods. The estimate of an appropriate washout period and the potential for confounding subsequent responses are, of course, problematic. Signs of poor tolerance or reaching a predetermined dose level will normally conclude this stage of development. Whenever possible, these studies involve concomitant assays of plasma drug levels.8
Table 16.1.1. Phase I drug development in healthy volunteers
Single dose Multiple dose Pharmacokinetics Clinical pharmacology Special populations
Adapted from Rosen et al."
Assessment of tolerability after single-dose administration Assessment of tolerability over likely dosing period Assess if preclinical studies predict kinetics and metabolites
Indirect assessment of efficacy; ensure adequate pharmacodynamic and pharmacokinetic relationships Children, elderly, and medical populations may display different tolerability/pharmacokinetics
To ensure that tolerability is not affected by administration of subsequent doses of the drug, multiple dose studies are also typical. The frequency of dosing depends on the predicted dosing regimen of the novel agent. Again, plasma drug levels are monitored whenever possible. Formal pharmacokinetic and drug metabolism studies are usually initiated after single- and multiple-dose tolerability studies. It is important to ensure that there is no dose-dependent tissue drug accumulation, and that the metabolic pathway and potential for major drug interactions are known.11
The value of equal numbers of placebo subjects in these safety studies should be emphasized. Incorrectly attributing adverse events to active treatment at this stage may haunt the rest of clinical development and even affect labeling for an approved product. Inclusion of adequate numbers of placebo subjects allows better discrimination of treatment-related effects from effects common to the site or the methodology of the study.
Efficacy assessment is first formally undertaken in phase II. Treatment assessment during this phase typically involves a mixture of objective measures (e.g., vaginal photoplethysmography) and self-report assessments (e.g., daily diaries, questionnaires). These measures are likely to be similar to, or serve as a pilot test for, the measures to be selected in the more extensive phase III studies. However, the primary objective of phase II efficacy studies is the identification of an effective dose range to be evaluated in phase III (dose setting). The range of doses in phase II studies is usually more limited than in phase I, since the
maximum tolerability of the drug has been established. A starting dose of 5-10% of the maximum tolerated dose may be initially studied. Additional information on drug tolerability is also obtained during this stage of clinical development, allowing calculation of potential benefit/risk ratio.11 Early phase II studies frequently serve as “proof-of-principle” to justify further research. Mid-phase II studies may establish a clear dose- response curve, and late phase II studies serve as “mini-pivotal” trials, testing issues of population definition, recruitment, and outcome measures. Preferably, both mid- and late phase II studies are parallel designs.
Crossover designs, in which each patient serves as her own control, are often used in early phase II trials. The major advantage of this design is the availability of within-subject comparisons and the associated reduction in patient variability and increased statistical power. Heiman et al.8 also note that various subject variables in women may be difficult to control (e.g., diet, body-mass index, alcohol intake, mood variability, menstrual cycle phase, precise hormonal status, and psychologic variables), but crossover designs minimize the potential error due to the variability from these factors. Potential problems are the likelihood of carryover and sequence effects from one treatment phase to the next and susceptibility to patient dropouts. These problems may be exacerbated if three or more treatment arms are included in the study design. The duration of treatment needed in each arm and potential for carryover effects are the major limitations, overall, of this type of design. Despite these limitations, crossover designs are frequently preferred in early phase II dose-setting studies. Different pharmacologic agents present different challenges in crossover designs. Those agents that are used “on-demand” and have a short pharmacokinetic and pharmacodynamic half-life (such as intranasal products) are best suited to this design. Hormone treatments, on the other hand, may have a pharmacodynamic half-life of weeks or months, and thus the washout period would be unwieldy.
When using a design that allows for “on-demand” treatment, it is important to recognize that outcome measures generally will reflect the patient’s attitude toward the treatment, rather than observations of specific episodes of treatment. Thus, the number of exposures to treatment needed to effect a stable change in a patient’s self-report may exceed the number of exposures required to demonstrate initial improvement. In effect, the half-life of an attitude is a major concern.12
Prior to regulatory approval, it is necessary to conduct two or more large-scale pivotal outcome studies in appropriately selected patient groups. Phase III trials are usually designed as multicenter, randomized, prospective studies with two or more dose levels of the study drug and a double-blind placebo control condition. In the field of female sexual dysfunction assessment, relatively long periods (e.g., 6 months) have been used for treatment arms in large-scale phase III trials. Since irreversible changes may occur during treatment periods of such long duration, crossover designs are typically not recommended for use in phase III studies.
Parallel designs are typically employed in phase III trials. In the simplest parallel design study, patients are randomly assigned to two or more parallel treatment arms for the duration of the study period (usually 4-6 months). Each patient is exposed to one treatment condition only, and comparisons are made between treatment groups at various time points. Treatment- induced changes are assessed by analysis of between-group differences after treatment. Results can also be analyzed relative to changes in baseline. Baseline assessments are necessary to ensure that the treatment groups are equivalent prior to randomization.11
The parallel treatment period, usually double-blind, is sometimes preceded by a single-blind placebo or “no treatment” run- in period to establish a reproducible baseline prior to randomization. This run-in period can also be used to screen out patients who are most susceptible to placebo effects or are unable to follow the protocol procedures. Following this stage, patients are assigned to the double-blind, randomized phase of the study. Large placebo effects (30-70% response rates) are commonly observed in clinical trials of female sexual dysfunction. The magnitude of placebo effects might be reduced through the use of single-blind, lead-in procedures with exclusion of identified placebo responders. However, this approach has not been widely employed in female sexual dysfunction trials to date, and thus it is unknown whether or how the exclusion of placebo responders after a run-in period would actually influence placebo effect sizes in the randomized trial period. Moreover, eliminating placebo responders may bias the selection of subjects to those who potentially have no responsiveness of any kind.
Phase IV studies are undertaken either during the approval process or subsequent to approval.11 One component involves “postmarketing surveillance” studies designed to expand the safety database through long-term tracking of patients on the active treatment. Special population studies may also be conducted during this phase. In general, phase IV studies are designed to increase understanding of the overall treatment profile in the target population(s). These studies may not include all of the controls (e.g., placebo-blinding and baseline assessment) utilized in phases II and III, although comparison treatments and special population groups are more likely to be included in the trial design during this phase.
Drug interaction studies
Women with sexual dysfunctions are likely in many cases to be receiving drug therapy for associated comorbidities (e.g., depression, menopausal symptoms, osteoporosis) (see Chapter 7.2). For this reason, it is important in the phase I and phase II studies that key drug interactions be examined. The interaction studies fall into two major categories: pharmacokinetic and pharmacodynamic interactions.
Pharmacokinetic drug interaction studies are designed to evaluate acute effects of the new agent on the plasma levels of other commonly used drugs that are concomitantly administered. Specific drug interaction studies may be required should the novel agent be known to induce or inhibit cytochrome- dependent liver metabolizing systems. Also important is the potential for pharmacodynamic interactions. Of particular relevance is the potential of novel drugs or hormonal preparations to influence the effects of other drugs. Alterations in blood pressure control in a woman with controlled hypertension and endocrine control in a diabetic woman, for example, would be undesirable features, and should be carefully evaluated during early phase I or phase II studies (see Chapter 7.3). Specific pharmacodynamic interactions relevant to the mechanism of action (e.g., centrally acting, peripherally acting) should be carefully assessed.
All clinical trials require a precise definition of which patients are and are not eligible for inclusion in the trial. The most important underlying principle guiding this definition is that the study population should represent the overall patient population for whom the treatment under investigation is intended. If the study population is truly representative of the intended treatment population, the results of a well-designed controlled trial are likely to predict the “real-world” effect. On the contrary, if the study population is too narrowly defined, the trial results may not generalize to the broader population. Therefore, when conducting “pivotal” phase III trials in female sexual dysfunction, it is important for the investigator to define a group of patients that will be as representative as possible of the intended patient population at large. This is accomplished by taking into account such factors as age, overall health status, concomitant medications, and the severity and duration of the disorder.
Careful patient selection using unambiguous inclusion and exclusion criteria should be sufficient to delineate a study population that is easily recognized by all those who assess the study results. The US Food and Drug Administration guidance document on clinical trials in female sexual dysfunction5 emphasizes the importance of specifying the hormonal status of the women (e.g., pre- or postmenopausal) in a clinical trial, as well as their specific sexual concern (e.g., female sexual arousal disorder, hypoactive sexual desire disorder). These factors should be controlled in any large-scale, clinical trial in women with female sexual dysfunction.
A third principle is that a logical and reasonable balance must be struck between the likely efficacy and safety of the treatment among enrolled patients and the “openness” of the entrance criteria. Specifically, the trial population should be sufficiently broad to represent the larger group of female patients who may eventually benefit from treatment, although it should not be so broad as to include patients who are clearly at direct high risk of injury from the study treatment or proce- dures.11 It should also be expected that treatments intended for broader clinical populations will be associated with broader variance in responsiveness, potentially jeopardizing the power of a study.12
Specific study population issues in female sexual dysfunction: representative patient population The phase III study population should include women with clearly defined subtypes of female sexual dysfunction, including or excluding hypoactive sexual desire disorder, sexual arousal disorder, orgasm disorder, or sexual pain disorder, depending on the desired indication. These sexual problems are frequently comorbid. Overlapping symptoms or comorbidities may weaken the diagnostic precision of the inclusion criteria for the study. Heiman et al.8 suggested that an alternative strategy is to select a group of women who meet diagnostic criteria for one disorder and have subsyndromal levels of another sexual disorder.
The phase III study population should provide a representative mixture of the various degrees of severity that can be expected in the target population. Ideally, this should be assessed by well-validated and treatment-sensitive instruments. These are reviewed in detail in Chapter 11.2. Stratified assignment of subjects to treatment by severity may be necessary to avoid a decrease in sensitivity due to the heterogenity of symptom severities between groups.
According to the recent review by Heiman et al.,8 the Female Sexual Function Index (FSFI)13 is the self-administered questionnaire of choice for clinical trials of female sexual dysfunction. The Food and Drug Administration guidance docu- ment,5 however, recommends the use of daily diary or event log measures to assess the frequency of satisfactory sexual events. Other female sexual dysfunction assessment tools are currently in development and advanced stages of testing.
Defining the disorder: inclusion and exclusion criteria
Phase III studies of female sexual dysfunction need to identify specific sexual function criteria for inclusion in the clinical trial.
Ideally, this assessment should be performed in a standardized way using a validated diagnostic tool. Questionnaire measures are not ideally suited for this, and a structured diagnostic interview is preferable. Several versions of this are in development, particularly for the assessment of hypoactive sexual desire disorder. A valid diagnosis requires that the patient demonstrate personal distress regarding her sexual function. This may be assessed by a validated instrument such as the Female Sexual Distress Scale.14 Since female sexual dysfunctions may be either primary (lifelong) or acquired, it is prudent to stipulate which situation applies in patient selection. Patients who were previously functional may have a higher probability of response to an intervention.
Hormonal status of the subjects must be addressed and may be defined by several factors: pre-, peri-, post-, or surgically menopausal status; the use of hormonal contraception; and the use of hormone therapy (including estrogens, progesterone, and androgens). Additionally, since the human teratogenic potential of a treatment may not be known, provisions for avoiding pregnancy during the study must be addressed.
Psychosexual and relational issues affect women’s sexuality and should be addressed in clinical trials. The quality of a patient’s relationship with her partner may greatly influence her response to interventions and satisfaction with her sexual function during treatment. The presence of sexual dysfunction in the partner may have serious confounding effects. The quality of the relationship should be addressed as part of the initial sexual history, and the presence of sexual dysfunction in the partner may be assessed by a validated instrument such as the Single Question Assessment of Erectile Dysfunction15 or a standardized, multidimensional measure such as the International Index of Erectile Function.16 The partner’s willingness to comply with the study requirements is another important variable to consider for the inclusion criteria.
Finally, inclusion criteria must specify the age range of participants, ability and willingness to provide informed consent, and ability to comply with study requirements. For example, if participants must attempt or engage in intercourse at regular intervals to determine the efficacy of a treatment, a trial may specify that participants be in a stable relationship with a person of the opposite sex. It is unknown whether including nonheterosexual women in a study might lead to higher variances in the study endpoints, and thus less sensitivity to detect treatment effects.
The major exclusion criteria in clinical trials of female sexual dysfunction typically include the following:
1. evidence of unresolved sexual trauma or abuse
2. female sexual dysfunction caused by untreated endocrine disease (e.g., hypopituitarism, hypothyroidism, diabetes mellitus)
3. pregnancy or breast-feeding
4. history of sensitivity to any of the ingredients in the test article
5. history or evidence of chronic or complicated urinary tract or vaginal infections within previous 12 months
6. history or evidence of pelvic inflammatory disease within previous 12 months
7. evidence of currently active sexually transmitted disease
8. history or evidence of currently active, moderate to severe vaginitis
9. history or evidence of current cervical dysplasia
10. history or evidence of current significant cervicitis as manifested by mucopurulent discharge from the cervix
11. history or evidence of significant gynecologic conditions such as uterine fibroids, vulvar vestibulitis, or vaginismus that may interfere with the patient’s ability to comply with study procedures
12. psychoses and bipolar disorder
13. use of neuroleptics or lithium, or any medications, herbal treatments, or dietary supplements intended to enhance sexual function within previous 3 months
14. history of myocardial infarction within previous 6 months
15. history or evidence of significant renal or hepatic disease within previous 6 months
16. significant central nervous system diseases within the last 6 months, such as stroke, spinal cord injury, or multiple sclerosis
17. any condition which, in the investigator’s opinion, would interfere with the patient’s ability to provide informed consent or comply with study instructions, or which might confound the interpretation of the study results
18. any condition which would endanger the participant if she participated in this trial.
Clinical trial endpoints and outcomes in female sexual dysfunction trials
Specific outcome measures or endpoints in clinical trials of female sexual dysfunction can be divided into four major categories or types. These usually include the following.
At present, several self-administered questionnaires are available, including the Brief Index of Sexual Functioning for Women17,18 and the Female Sexual Function Index.13 These questionnaires assess sexual function across a number of domains (e.g., sexual desire, arousal, orgasm, satisfaction) and measure average responses over a specified time period (e.g., 4 weeks). Scoring algorithms are provided for computing average responses in each of the specified sexual function domains. Both have demonstrated good reliability and validity. Chapter 11.2 describes these measures and other self-administered questionnaires in more detail.
Event logs and daily diaries
A variety of event logs and daily diaries have been used in clinical trials of male and female sexual dysfunction. These diaries or event logs are completed immediately after each episode of sexual activity and are designed to assess sexual function and satisfaction during each sexual episode. The Female Sexual Encounter Profile12 has been employed in many studies. Several other event logs and diaries are under development. Data obtained from the sexual event log are typically regarded as primary or secondary endpoints from a regulatory perspective. Event logs are currently favored by the Food and Drug Administration to assess “satisfactory sexual events”.5
Although interview protocols have been developed for in-depth assessment of sexual function,19 these methods have not been used extensively in clinical trials of male or female sexual dysfunction. Other interview tools currently under development may result in greater diagnostic precision for female sexual dysfunction, and greater specificity in the inclusion criteria.
Current physiologic measures, such as vaginal photoplethysmography, lack adequate standardization or clinical validation, and are not widely used in multicenter, large-scale treatment studies. These measures may be useful, however, in early phase II studies, or in mechanistic or early dose-finding studies. However, the predictive value of these studies from the laboratory setting to the at-home phase with self-report measures is not well established.
Primary and secondary endpoints should be specified in advance of the study. Secondary endpoints typically include other domains of sexual function (e.g., sexual desire, orgasmic satisfaction) and/or quality-of-life measures (e.g., relationship satisfaction, overall satisfaction). These measures have been recently described in detail elsewhere.8
Clinical and statistical significance
Clinical trials are typically powered to show a statistical difference between the active drug and placebo on the primary endpoint at the 0.05 level of significance or higher. When a large number of subjects are included in each arm of the study (e.g., 100 per treatment arm), a high degree of statistical significance may be obtained (e.g., p < 0.001) with relatively small mean differences between groups on the primary study endpoint. For example, a 1- or 2-point change on a self-report (self- administered questionnaire) scale might be statistically significant in a well-powered study, but may nonetheless represent a very slight difference or change in actual behavior. In the absence of clear-cut norms for determining clinical relevance, the interpretation of clinically meaningful change presents a serious challenge in the field. At present, the interpretation of clinically relevant change is highly variable from one study to another, and no published standards or norms are available. This is an important scientific and regulatory question that warrants attention.
Sexual dysfunction in women is characterized by multiple, diverse etiologic determinants, a high degree of overlap between disorders, and a high frequency of comorbidity with mood disorders and other medical and mental health disorders. These factors complicate both the design of clinical trials and the interpretation of treatment effects. Although the elimination of women with comorbid disorders may be an appealing means of controlling for confounding variables, this can easily limit the generalizability of the findings. In designing a clinical trial for female sexual dysfunction, investigators need to consider the multiple etiologies and determinants of female sexual function, including hormonal status, psychologic factors, and issues pertaining to the relationship with the sexual partner. Each of these domains of influence affects the selection of subjects and assessment measures for a trial, and may influence the baseline levels of key endpoints. These issues should be carefully considered in planning the statistical analytic strategy for the trial.20
Diverse influences on sexual function in women require assessing treatment outcomes and satisfaction by multidimensional measures with sufficient breadth and sensitivity of measurement.4 Some controversy surrounds methodological approaches to the assessment of treatment outcomes in clinical trials of female sexual dysfunction. Although the Food and Drug Administration has recommended the use of daily diary or event log measures to assess “satisfactory sexual events”,5 this position has recently been challenged in a consensus white paper on selection of endpoints in clinical trials of female sexual dysfunction: “Counting events is simple; what determines a woman’s satisfaction or success with an event is less clear. Moreover, the judgment is entirely subjective, more nuanced and best captured by a self-administered questionnaire.”6
We have emphasized the role of psychologic and relationship factors throughout as etiologic and maintaining factors in female sexual dysfunction. These factors very often are unaddressed or not controlled for in clinical trials of female sexual dysfunction, but are typically assumed to be operating in a constant state or level throughout the study. In fact, changes in sexual function in the woman are likely to initiate responses or changes in the couple’s relationship, as well as in the woman’s self-image or sexual self-schema.21 These psychologic changes are likely to be determining factors in whether or not treatment changes are maintained during and after a clinical trial of female sexual dysfunction. Relationship factors and changes in the woman’s partner may mediate her response in important ways. Typically, clinical trials in female sexual dysfunction have neither controlled for nor assessed these factors in the design or outcomes components of the trial.
In summary, despite the high prevalence of sexual problems in women, clinical trials in female sexual dysfunction have lagged significantly behind comparable treatment studies in male sexual dysfunction. The study design and choice of outcome measures have created opportunities and challenges for clinical investigators in this area. The lack of a common pathophysiologic model or clear-cut consensus regarding the diagnostic classification of female sexual dysfunction, and the determination of inclusion and exclusion criteria for clinical trials, has made it difficult to develop consistent standards for clinical trials. Finally, the difficulty in defining a treatment responder and the related need to establish clinically meaningful change criteria have been addressed in this chapter. Despite these current areas of challenge, major advances are anticipated in the clinical trial aspects of female sexual dysfunction diagnosis and treatment in the near future. This should lead to the eventual development of safe and effective long-term treatments.
The authors of this chapter acknowledge in particular the contributions of the First International Consultation on Erectile Dysfunction and the Second International Consultation on Sexual Medicine committees, and for permission received from the Consultations to adapt or reprint source material from their proceedings.
1. Tiefer L, Hall M, Tavris C. Beyond dysfunction: a new view of women’s sexual problems. J Sex Marital Ther 2002; 23: 225-32.
2. Segraves RT. Historical and international context of the nosology of female sexual dysfunction. JSxMûïÙalTher 2001; 27: 205-7.
3. Basson R. A model of women’s sexual arousal. J Sex Marital Ther 2002; 28: 1-10.
4. Meston CM, Derogatis LR. Validated instruments for assessing female sexual function. J Sex Marital Ther 2002; 28: 155-64.
5. US Food and Drug Administration, Center for Drug Evaluation and Research. Female sexual dysfunction: clinical development of drug products for treatment. Rockville, MD: US Department of Health and Human Services, 2000.
6. Althof SE, Rosen RC, DeRogatis L, Corty E, Quirk E Symonds T. Outcome measurement in female sexual dysfunction clinical trials: review and recommendations.
7. Bancroft J, Loftus J, Long JS. Distress about sex: a national survey of women in heterosexual relationships. Arch Sex Behav 2003: 32: 193-208.
8. Heiman JR, Guess MK, Connell K et al. Standards for clinical trials in sexual dysfunctions of women: research designs and outcomes assessment. In TF Lue, R Basson, R Rosen et al., eds. Sexual Medicine: Sexual Dysfunctions in Men and Women. 2nd International Consultation on Sexual Dysfunctions. Paris: Editions 21, 2004.
9. Patton MQ. Qualitative Research and Evaluation Methods, 3rd edn. Thousand Oaks: Sage, 2002.
10. Basson R, Berman J, Burnett Aetal. Report of the international consensus development conference on female sexual dysfunction: definitions and classifications. J Urol 2000; 163: 888-93.
11. Rosen R, Bennett A, Ferguson D et al. Standards for clinical trials in erectile dysfunction: research designs and outcomes assessment. In A Jardin, G Wagner, S Khoury et al., eds. Erectile Dysfunction. Plymouth: Health Publications, 2000.
12. Ferguson DM. Clinical trial development in female sexual dysfunction. JSxMaiâLThei 2002; 28: 77-83.
13. Rosen RC, Brown C, Heiman Jetal. The Female Sexual Function Index (FSFI): a multidimensional self-report instrument for the assessment of female sexual function. J Sex Marital Ther 2000; 26: 191-208.
14. Derogatis LR, Rosen R, Leiblum Setal. The Female Sexual Distress Scale (FSDS): initial validation of a standardized scale for assessment of sexually relevant personal distress in women. J Sex Marital Ther 2002; 28: 317-30.
15. Derby CA, Araujo AB, Johannes CB et al. Measurement of erectile dysfunction in population-based studies: the use of a single question self-assessment in the Massachusetts Male Aging Study.
16. Rosen RC, Riley A, Wagner G et al. The International Index of Erectile Function (IIEF): a multidimensional scale for assessment of erectile dysfunction. Urology 1997; 49: 822-30.
17. Taylor JE, Rosen RC, Leiblum SR. Self-report assessment of female sexual dysfunction: psychometric evaluation of the Brief Index of Sexual Functioning for Women. ArchSxBehav 1994; 23: 627-43.
18. Mazer NA, Leiblum SR, Rosen RC. The Brief Index of Sexual Functioning for Women (BISF-W). A new scoring algorithm and comparison of normative and surgically menopausal populations. Menopause 2000; 7: 350-63.
19. Derogatis LR. The Derogatis Interview for Sexual Functioning (DISF/DISF-SR): an introductory report. J Sex Marital Ther 1997; 23: 291-304.
20. Islam A, Mitchel J, Hays Jetal. Challenges in conducting multicenter clinical trials in female sexual dysfunction: baseline differences between study populations. J Sex Marital Ther 2001; 27: 525-30.
21. Cyranowski JM, Andersen BL. Schemas, sexuality, and romantic attachment. JPersSocPsychol 1998; 74: 1364-79.